Public opinion polls in Iraq since 2003 have been crucial to understanding the war-torn country. ABC News relied on polls for reporting that was awarded two Emmys -- the first to mention public opinion polling.
But two researchers looking at Iraq polling data in 2011 found alarming patterns that they said suggested some results may have been fabricated by people in Iraq. They wrote a paper describing their findings and sent it to the U.S. company in charge of the data collection, D3 Systems, and to some of the clients who used the data, including the pollsters who had worked with ABC at the time.
But instead of opening a discussion of the researchers' findings, D3's attorneys issued a cease-and-desist letter, threatening to sue if the paper was distributed or published.
Now, four years later, Michael Spagat, a professor at London's Royal Holloway College, has posted the original research paper and has disclosed the cease-and-desist letter from D3's law firm, Lev and Berlin. (Read the letter here.)
Spagat and his co-author, pollster Steve Koczela -- who is no longer associated with the project -- intended to publish the paper in a professional journal, but said he first shared it with D3, and then the clients, to give them a chance to respond.
The cease-and-desist letter alleges the researchers wanted to harm D3’s business by maligning the company to its clients. It also says the company's own research showed problems found by Spagat and Koczela didn't actually exist. The letter didn't include specifics.
Matthew Warshaw, vice president of D3 Systems, told The Huffington Post on Friday night that the letter stemmed from the way the researchers handled their findings.
"The only reason we went the legal route was because [Spagat and Koczela] sent their work to our clients first, before they sent it to us," Warshaw said. "Why did it go to our clients first if the concern was to open a discussion about data fabrication issues?"
Representatives of Langer Research Associates, the company later formed that includes the pollsters who worked for ABC, and Lev & Berlin didn't immediately respond to HuffPost's requests for comment after business hours on Friday.
Polling the Iraqi population during a war was extremely difficult. Interviewers were instructed to go door to door with questions.
In surveys done this way, particularly in war zones, the potential exists for interviewers or their supervisors to make up data instead of entering a potentially dangerous area. Indeed, this is a known problem with surveying in unsafe circumstances.
Other polls in Iraq at the time, notably those trying to get an accurate death count, were plagued by the difficult circumstances and claims of potential falsification.
Issues of data falsification are controversial in polling. Any false data undermines the quality and reliability of a survey, and pollsters rely on their reputations for producing high-quality, reliable data in order to get work.
Usually, these debates play out in the open. The Pew Research Center recently released a report refuting claims of fraud in some of its data.